We investigate the effect of universal health insurance coverage on health outcomes and the use of health services by exploiting a natural experiment that changes the insurance status of most Americans at age 65, that is, eligibility for the U S Medicare program Since instance status is not randomly assigned, correctly identifying a causal relationship between insurance status and health is notoriously difficult This study will employ panel data from the Health and Retirement Study in the U S to estimate the effect of turning 65 and becoming Medicare eligible on health status and the user of health services The Clinton Administration's 1998 introduction of legislation seeking to expand Medicare coverage to many individuals in the 55-to-64 year age group (a group often referred to as the "near elderly") has since spawned considerable interest among policy analysts who have sought to document the insurance, health and employment status of this age group The enclosed study explores the possible benefits of such a proposal by examining the effect of obtaining Medicare coverage on the health of previously uninsured individuals in the near elderly group Many researchers both in the U S and abroad have documented that socioeconomic differences in health tend to be largest in middle age (prior to age 65), and decline at older ages The proposed study will also fill a large gap in this literature on converging socioeconomic inequalities in health over the lifecycle by exploring the role of universal health I insurance coverage in producing these patterns In particular, we will compare differences in health just before and after the age of 65 in the U S to differences between comparable ages in Canada, where Medicare is a universal program and eligibility is not conditioned on age.